Enterprise deployment of artificial intelligence (AI) is positioned for tremendous growth.
Artificial intelligence is set to change the business world by improving predictive analytics, sales forecasting, customer needs, process automation and security systems.
IBM’s Global AI Adoption Index revealed that a third of those surveyed will be investing in AI skills and solutions over the next 12 months.
More expansive use of AI democratizes AI, providing access to insights to more people – technologists and non-technologists alike. The latter group might include people in leadership, sales, finance, human resources and operations. This is where AI will shine, empowering business teams to make AI-driven decisions.
Imagine: business teams do not have to know how to code or be schooled in the intricacies of AI’s backend. Instead, they will use AI like you and I use a mobile phone for efficiency (if we’re running late, we merely send a text notifying the other person), access information faster (if we’re in the grocery store and need a recipe, we look it up), make better decisions (GPS gives us the fastest route).
Just as mobile technology works without us understanding complex circuitry, algorithms or software, the democratization of AI across enterprises will be integrated in much the same way.
So, what will hold AI back and how will AI help enterprise companies gain traction?
Artificial intelligence deployment approach | Source: IBM
Obstacle #1: Data in disarray. Data that does not provide a complete picture and single version of truth because of data silos and various data formats within an organization.
Opportunity: Employing a data fabric. Using a data fabric to help organizations use data more effectively and get the right data to users regardless of where it is stored. One significant advantage of a data fabric is that data governance rules may be automatically set for compliance.
Having one information structure to garner insights and analytics from, integrating security to protect sensitive data and establishing a framework for implementing trustworthy AI positions AI as part of the business strategy, not solely an IT strategy so that AI directly impacts business operations.
It all comes back to connecting data with business drivers and a data fabric helps accomplish this. It is what I call “point-to-point” thinking – knowing the business imperatives, business drivers, the different levels of raw data, who is consuming the data, who will have access to the data, and why the data is important in decision-making and then, the big payoff with AI, how it will elevate experiences: customer experience, workforce experience, supply chain experience, strategic partner experience, community experience. In “point-to-point” thinking we don’t hoard data, but share it – securely.
Obstacle #2: Varied skill levels. A lack of AI technical skills across the enterprise and a reliable, open platform to bring AI to more people.
Opportunity: Creating a bridge to AI for people within the enterprise. Palantir for IBM Cloud Pak for Data is one of the great innovations of our time because it doesn’t require coding skills. People in non-technical roles can go from raw data to data insights quickly using application templates (think of all the designs being produced with minimal design experience because of apps like Adobe Photoshop and Canva). This is truly the path to democratizing AI.
People can now use AI to make better decisions in real time and improve business outcomes. These teams include sales and marketing, manufacturing operations, campaign managers, branch managers, franchise operators, human resources, among others.
An example: a customer walks into their regional bank. The banking professional greets the customer, invites them to sit down and pulls up their profile. They see, not only account information, but a 360-degree view of the person sitting across from them. Through a data fabric, non-tabular visualizations gathered from previously siloed data originating from different systems provides an AI-infused perspective.
This might include two algorithmically recommended customer offers inspired by marketing analyst data and intelligent customer segmentation and campaign propensity scoring powered by Watson models.
Going further, feedback from the customer can then be entered and that data influences future offers because it goes right back into IBM’s data and AI platform. IBM Cloud Pak for Data, which helps to simplify data management and protect sensitive data by establishing a framework for implementing trustworthy AI.
Obstacle #3: Solving for the wrong “x.” In hundreds of conversations I’ve had with enterprise leaders over the years about AI, one common failure I see not identifying the right problem or identifying use cases that will yield high return from AI.
Opportunity: Clearly articulating the problem to be solved. With AI, we are talking about a machine making reasonable conclusions based on data. Better defining the problem is akin to asking better questions.
Imagine the difference if you were in a store and asked someone if they sold products. The question is too vague to expect a meaningful answer. Ask where the tomatoes are and you get a clear answer. Both are valid questions, but one is more focused. That’s how defining the problem should be (this is not just for AI purposes; I devote a lot of space in my book, Ascend Your Startup, to defining the customer problem because I believe building the wrong solutions plagues many companies)
In an interview with famed Mount Everest climber George Mallory, a reporter asked him why he wanted to climb the formidable mountain. His answer: “Because it’s there.” AI is very much the same thing. It has obstacles, yet it has the allure of opportunity and of making measurable progress.
Here are the big three takeaways for enterprise companies:
• Use a data fabric. Information is powerful – and it exists! Don’t let siloed data and inconsistent data formats hold people back from making better decisions.
• Give people what they need to succeed in their jobs. Tools such as low code/no code enable business users to rapidly leverage data and apply AI in their decision making.
• Go back to square one and define the problem. Solving for “x” without fully understanding “x” wastes precious time, causes unnecessary frustration and marginalizes the experience for everyone involved.
In a Forbes article on the topic of AI, author Manas Agrawal writes, “With rapid learning and adoption, AI is no longer a crystal ball technology but something that humans now interact with in nearly every sphere of life.”
In a very short time, we won’t be talking about AI adoption as people see it as part of doing business and part of making life more efficient. AI then will shift to being part of an enterprise’s business strategy, delivering value for non-technical people working in many different areas like customer experience, brand differentiation, HR, research and development, management and sales.
This is what the democratization of AI looks like at the crossroad of technology and humanity to improve outcomes for people leading successful enterprise businesses.